STATUS REPORT OF THE BERLIN ENERGY RECOVERY LINACPROJECT bERLinPro∗
M. Abo-Bakr†, W. Anders, A. Büchel, K. Bürkmann-Gehrlein, A. Bundels,Y. Bergmann, P. Echevarria, A. Frahm, H.-W. Glock, F. Glöckner, F. Göbel, B. Hall, S. Heling,
H.-G. Hoberg, A. Jankowiak, C. Kalus, T. Kamps, G. Klemz, J. Knobloch, J. Kolbe, G. Kourkafas,J. Kühn, B. Kuske, J. Kuszynski, A. Matveenko, M. McAteer, A. Meseck, R. Müller, A. Neumann,
N. Ohm-Krafft, K. Ott, E. Panofski, L. Pichl, F. Pflocksch, J. Rahn, M. Schmeißer, O. Schüler,M. Schuster, J. Ullrich, A. Ushakov, J. Völker,
Helmholtz-Zentrum Berlin für Materialien und Energie GmbH, Berlin, Germany
AbstractThe Helmholtz-Zentrum Berlin is constructing the En-
ergy Recovery Linac Prototype bERLinPro, a demonstrationfacility for the science and technology of ERLs for futurelight source applications. bERLinPro is designed to acceler-ate a high current (100 mA, 50 MeV), high brilliance (norm.emittance below 1 mm mrad) CW electron beam. We reporton the last year’s progress, including the comissioning ofthe gun module as the first SRF component to be installedin bERLinPro.
INTRODUCTIONbERLinPro [1] is an Energy Recovery Linac Prototype,
currently under construction at the Helmholtz-ZentrumBerlin (HZB), Germany. Application of superconductingradio frequency (SRF) systems will allow to accelerate cur-rents at storage ring levels. The layout is shown in Fig. 1,the project’s basic set of parameters is listed in Table 1. ThebERLinPro injector, consisting of an photoinjector cavity(1.4 cell), followed by a Booster module containing threeSRF cavities (2 cells), generates a high brilliant beam with anenergy of 6.5 MeV. This beam is merged into the main linacsection by means of a dogleg chicane and then accelerated inthe three SRF cavities (7 cells) Linac to 50 MeV. With a race-track magnetic lattice, the beam is recirculated for energyrecovery and then sent into a 650 kW beam dump. Space isprovided in the return arc to install future experi- ments orinsertion devices to demonstrate the potential of ERLs foruser applications. Major construction in the building wascompleted in 2017 so that machine component installationhas begun. The accelerator installation is planned in twostages: the first stage, called "Banana", includes the entirelow energy beam path from the gun to the high power beamdump, with a 5 mA gun (Gun1) and the Booster as well as adiagnostics line. The installation of the "Banana" is ongo-ing: after placing and aligning all girders and magnets in thebeginning of 2017, rf and cryogenic installations are nearingcompletion. The Banana vacuum system will be assembledin the second half of this year, followed by the installation ofthe SRF modules (gun and booster). In the second project
∗ Work supported by the German Bundesministerium für Bildung undForschung, Land Berlin and grants of Helmholtz Association† [email protected]
Table 1: bERLinPro’s Main Target Parameters
parameter value
maximum beam energy / MeV 50maximum average current / mA 100normalized emittance / µm rad 1.0bunch length / ps 2.0 (0.1)rf freq. & max. rep. rate / GHz 1.3maximum losses < 10−5
Figure 1: Basic bERLinPro layout. The green highlightedbeam path is indicating the "Banana".
stage the installation and commissioning of the high currentelectron source, the linac module and the recirculation loopis planned to demonstrate efficient energy recovery with thefull current, 50 MeV beam.
After the first SRF gun test at HZB in 2011 [2,3], in 2017the second one took place in GunLab, a dedicated gun testlaboratory [4]. Results of this run together with an overviewof last year’s progress of the various subproject groups isprovided in this paper as well as update on the further projectplanning.
GUN COMMISSIONINGSince setup of the SRF Gun and diagnostics beamline
[5] the goal was to complete technical commissioning ofthe diagnostics beamline and to start beam operation withGun1, initially with metallic Cu photocathodes, later withmulti-alkali CsK2Sb photocathodes. We encountered severaltechnical problems during the pre-beam check out phase.We were able to fix most problems in the warm part of thediagnostics beamline but were left with a short circuit in the
current leads of the SC solenoid in the cold mass of the SRFgun module. This could not be fixed jeopardizing all beamdynamics related measurements planned for commissioning.
The drive laser beamline could be completed and is nowserving both UV and green output wavelengths of the drivelaser. For commissioning of Gun1 with Cu cathode UV at260 nm can be generated from the drive laser fundamentalwavelength, for the CsK2Sb photocathode green laser pulsesat 515 nm can be send to the cathode.
The goals related to photocathode R&D were preparationof Cu and CsK2Sb photocathodes for the commissioning andthe establishment of the particle-free UHV transfer chain [6]for these photocathodes. This transfer chain connects thepreparation system which is located in a separate buildingthan the SRF gun roughly 1000 m away. The transport chainis now fully operational and could be successfully testedwith several Cu photocathodes.
Of particular interest during the last year was the setup andcommissioning of two systems, monitoring the photocathodeposition inside the SRF gun cavity [7]. The position ofthe cathode surface with respect to the cavity backwall isstrongly influencing the field distribution in the cavity andon the normal conducting cathode surface. It is thus definingboth, the accelerating to focusing field ratio, an importantparameter for the beam dynamics, and the ohmic losses inthe cathode potentially leading to significant heating.
For this reason the cathode plug is movable and the posi-tion will be monitored with two independent systems. Oneis based on a laser distance meter, the other on a pack ofthree capacitive distance sensors.
We were able to prepare and transfer one Cu cathode intothe SRF gun cavity. First beam could be generated and asmall-scale program aimed at the characterization of the SRFgun cavity and the cathode (quantum efficiency mapping,dark current) with beam was started [8]. Then a multi-alkaliCsK2Sb photocathode with a quantum efficiency (QE) of16.8% at 515 nm was prepared and transported into the SRFgun transfer system. The QE after transport was 5.3% at515 nm after six days in the gun transfer chamber, limited byresidual gas. During transfer into the SRF gun cavity we lostthe photocathode plug due to a technical failure of the plugholding mechanism. Operation was stopped and concluded.We are now in the process to repair the SRF gun.
SRF SYSTEMSGUN1.X: after installing GUN1.0 into the gun module in
the first half of 2017, it has been commissioned and operatedwithin the GunLab framework in the second half of 2017(see [8]). In parallel the production of a second identicalsubstitute gun (GUN1.1) was running, see [9] for details.This cavity will replace GUN1.0 and be installed into thegun module this year.
Booster: while still most activity of 2017 concentratedon the setup and commissioning of the first SRF photo-injector module, also the preparation to assemble the Boostermodule was intensified. Figure 2 displays a comparison
Figure 2: Comparison of vertical test at JLab to Q0(Eacc)
data measured in horizontal test setup at HZB in HoBiCaT(see insert).
between quality factor Q0 versus mean accelerating fieldEacc measured with the final vertical test at JLab [10] andhorizontal data at HZB after controlled venting and pumpingof the cavities in the cleanroom. They are now ready forcold string assembly. This can start, once the high powercouplers are in house and conditioned in a dedicated testbox [11]. As the vendor had issues with the electro-polishingof the solid Copper-made inner conductors of the cold part,delivery will be now expected for early fall this year.
WARM SYSTEMS
Magnets: the gaps of the low energy path magnets areopen since their installation in Q I /2017, and will be closedby the BINP colleagues after installation and baking out ofthe vacuum-chambers in December 2018. At this time alsoall magnet-gaps of the recirculator will be opened to preparefor installation of the vacuum chamber.
Vacuum system: in March 2018 the first vacuum compo-nents for the "banana" have been delivered. Due to bi-metalcomponents of low quality nearly all the already weldedflanges hat to be exchanged. This caused a delay of deliveryand installation schedules. The first half of the low energy,"banana" vacuum components will be installed in July 2018followed by the second installation phase in October 2018.This work will be done under responsibility of the manufac-turer. Nevertheless the preparation of the accelerator hall aswell as the provision and operation of the clean room tentsis on HZB’s own authority. Nearly all the cables and wiresfor the diagnostic and vacuum components are installed andassembling is ongoing. All the electrical, cooling-water andpressed-air infrastructure for the warm system componentsis installed in the accelerator hall.
Beam dumps: in December 2017 the main dump wasdeliverer and is now installed in the accelerator hall, as thefirst vacuum component of the "banana". Vacuum and watertightness were tested at the manufacturer site. Cleaningand backing out of the big copper hat will take place beforeinstallation of the other low energy beam path components.
02 Photon Sources and Electron AcceleratorsA18 Energy Recovery Linacs
Figure 3: Feedboxes and cryolines in the bERLinPro accel-erator hall. In the background the injector line and the endof the second recirculator arc can be seen.
Beam diagnostics: all main components of the diagnos-tics are delivered and tested in house. The strip-line socketsfor the low energy path were measured before finial weld-ing to the vacuum chambers. DCCTs and FCT were testedfor different operation modes and the limits in stability andaccuracy were examined. The beam loss monitor sensorswithstood the radiation exposure tests and are in production.
Machine Protection System (MPS): the MPS will pre-vent severe machine damages caused by losses of the highpower beam. It is specified to reach a reaction time of ap-proximately 2 − 3 µs between receiving a diagnostic inputsignal and sending out a shut off output signal. The proto-type of a scalable and distributed, FPGA based MPS withEtherCAT communication link to the EPICS based controlsystem has been tested successfully. The production is onthe way in order to provide the MPS in time.
RF AND CRYOGENICSCable, wave guide and controls installations are progress-
ing well and will be timely finished. All gun and boostertransmitter parts are in house. The first one is on power andcurrently tested, the others are put on place in the bERLin-Pro building. In April 2018 the call for tender for the linacsolid state amplifiers (4 with 15 kW each) was started.For the cryogenics system all parts in house: cold com-
pressors, warm vacuum pumps and the module feed boxes,all flexible and rigid cryo lines are installed, see Fig. 3. InFebruary the existing Helium liquifier L700 moved to thebERLinPro infrastructure hall.
RADIATION SAFETYBesides the radioactivation of magnets and vacuum sys-
tem [12] also that of the cooling water was considered. Sincethe heat exchanger are outside the accelerator enclosure thedose rate around the water tubes was determined as well asthe activation concentration of the radionucleii with longerhalf lives. The later one gives the decline and thus storagetimes in case of a cooling water leakage.
In a first step we calculate the radionucleii and the doserates with Fluka [13, 14] using a model with a 2 m longaluminum vacuum tube of elliptical shape including twocooling water tubes (diameter 1 cm) on both sides of the
Table 2: Activitation Concentration, with the Limits of theGerman Radiation Protection Ordinance in the Last Column
nuclide A(t) T1/2 C/ Bqcm3 CL / Bq
cm3
31H 1.159E+04 12.323 a 72.34 100074Be 5.028E+04 53.29 d 313.82 30146 C 87.7 5730 a 0.54 80
electron beam. A beam of 100 µA loss current is hitting theside of the vacuum tube with an angle of 20 mrad.
With the same Fluka run also the production rate ÛN+ ofthe radionulceii has been calculated. From that the activitiesof the radionucleii was calculated (irradiation pattern: 8 hbeam, 16 h decline, for 365 periods), using the activationequation, see e.g. [12].
By scaling these results with the full accelerator dimen-sions we get the activation concentrations for the completewater volume of the cooling circuit shown in Table 2. Whilethe activation concentrations of Tritium and Carbon are farbelow the limit, the one of Beryllium is about one order ofmagnitude above the limit. Thus, the water has to be storedand controlled for a decline time of 10 half-lives before itcould be given to effluents system.
In a last step from the activation concentrations the doserates from the radionucleii were calculated, including alsoradionucleii with short half-lives. The dose rate is < 1 µSv/heven close to the water tubes, thus not significant.
PROJECT TIME LINEDue to technical and delivery issues, as well the competing
BESSY VSR project, slippage of the bERLinPro schedulehas been unavoidable. Prioritized BESSY VSR developmentcoincides with the planned assembly times of the high cur-rent gun and the main linac fully, straining available staffingand the infrastructure. Thus these components are currentlyon hold. To improve the situation, planning is currently un-der way to temporarily install one of the MESA modulesfrom the Universität Mainz ERL project [15]. This optionwould allow for a mA class operation of bERLinPro, at nearlythe project target energy and to demonstrate energy recovery.Table 3 provides the time line, including the MESA option.No times for the high current operation of bERLinPro witha dedicated linac module are given.
Table 3: bERLinPro’s Updated Time Line
Q1/2017 building ready for machine installation,start of cryo system installation & comm.
Q3/2017 first electrons from GUN-1 @GunLab2018 Refurbishment of GUN1
Q1/2019 start SRF operation GUN-1 @BananaQ2/2019 first electrons in Banana
2020 first electrons with MESA optionGUN-1, Booster, MESA module & recirc.
REFERENCES[1] M. Abo-Bakr et al., Status Report of the Berlin Energy Re-
covery Linac Project, Proc. of IPAC 2017, Copenhagen, Den-mark, 2017.
[2] T. Kamps et al., First Demonstration of Electron Beam Gen-eration and Characterization with an All SuperconductingRadio-frequency (SRF) Photoinjector, Proc. of IPAC 2011,San-Sebastian, 2011.
[3] A. Neumann et al., First Characterization of a Fully Super-conducting RF Photoinjector Cavity, Proc. of IPAC 2011,San-Sebastian, 2011.
[4] J. Völker et al., Introducing GunLab - A Compact Test Fa-cility for SRF Photoinjcetors, Proc. of IPAC 2014, Dresden,Germany, J2014.
[5] T. Kamps, et al., Setup and Status of an SRF Photoinjectorfor Energy-Recovery Linac Application, Proc. of IPAC 2017,Copenhagen, Denmark, 2017.
[6] J. Kühn, et al., UHV Photocathode Plug Transfer Chain for thebERLinPro SRF-Photoinjector, Proc. of IPAC 2017, Copen-hagen, Denmark, 2017.
[7] J. Kühn, et al., A Cu photocathode for the Superconduct-ing RF photoinjector of bERLinPro, Proc. of IPAC 2018,Vancouver, Canada, 2018.
[8] A. Neumann, et al., The bERLinPro SRF Photoinjector Sys-tem - From First RF Commissioning to First Beam, Proc. ofIPAC 2018, Vancouver, Canada, 2018.
[9] H.-W. Glock, et al., Preparation and Testing of the bERLinProGun 1.1 Cavity, Proc. of IPAC 2018, Vancouver, Canada,2018.
[10] A. Burrill et al., First Test Results of the bERLinPro 2-cellBooster Cavities, Proc. of IPAC 2015, Richmond, VA, USA,2015.
[11] B.D.S. Hall et al., 120kW RF Power Input Couplers forbERLinPro, Proc. of IPAC 2017, Copenhagen, Denmark,2017.
[12] M. Helmecke, K. Ott, Activation Calculations with FLUKAfor the bERLinPro Vacuum System, Proc. of RADSYNCH2013, Brookhaven, NY, USA (2013).
[13] G. Battistoni, et al., Proc. of the Hadronic Shower SimulationWorkshop 2006, Fermilab 6–8 September 2006, M.Albrow,R. Raja eds., AIP Conference Proceeding 896, 31-49, (2007).
[14] A. Ferrari, P.R. Sala, A. Fassò, and J. Ranft, FLUKA: amulti-particle transport code, CERN Yellow Report (2005).
[15] F. Hug et al., MESA - an ERL Project for Particle PhysicsExperiments, Proc. of Linac 2016, East Lansing, USA, 2016.
